paper-with-me

Papers

Masked Language Modeling and the Distributional Hypothesis: Order Word Matters Pre-training for Little

2021-04-14 · EMNLP 2021 11 · Koustuv Sinha, Robin Jia, Dieuwke Hupkes, Joelle Pineau, Adina Williams, Douwe Kiela

A possible explanation for the impressive performance of masked language model (MLM) pre-training is that such models have learned to represent the syntactic structures prevalent in classical NLP pipelines. In this paper, we propose a different explanation: MLMs succeed on downstream tasks almost entirely due to their ability to model higher-order word co-occurrence statistics. To demonstrate this, we pre-train MLMs on sentences with randomly shuffled word order, and show that these models still achieve high accuracy after fine-tuning on many downstream tasks -- including on tasks specifically designed to be challenging for models that ignore word order. Our models perform surprisingly well according to some parametric syntactic probes, indicating possible deficiencies in how we test representations for syntactic information. Overall, our results show that purely distributional information largely explains the success of pre-training, and underscore the importance of curating challenging evaluation datasets that require deeper linguistic knowledge.

📄 PDF Abstract BibTeX arXiv:2104.06644

Code (0)

등록된 구현이 없습니다.

Tasks

Language ModelingLanguage ModellingMasked Language Modeling

Similar Papers 제목 키워드 기반

The Distributional Hypothesis Does Not Fully Explain the Benefits of Masked Language Model Pretraining

2023-10-25 · Ting-Rui Chiang, Dani Yogatama

We analyze the masked language modeling pretraining objective function from the perspective of the distributional hypothesis. We investigate whether better sample efficiency and the better generalization capability of mo…

Language ModelingLanguage ModellingMasked Language ModelingSemantic Similarity+1

Can Eye Movement Data Be Used As Ground Truth For Word Embeddings Evaluation?

2018-04-23 · Amir Bakarov

In recent years a certain success in the task of modeling lexical semantics was obtained with distributional semantic models. Nevertheless, the scientific community is still unaware what is the most reliable evaluation m…

Embeddings EvaluationSentenceWord EmbeddingsWord Similarity

Masked Language Modeling Becomes Conditional Density Estimation for Tabular Data Synthesis

2024-05-31 · SeungHwan An, Gyeongdong Woo, Jaesung Lim, Changhyun Kim 외

In this paper, our goal is to generate synthetic data for heterogeneous (mixed-type) tabular datasets with high machine learning utility (MLu). Since the MLu performance depends on accurately approximating the conditiona…

Density EstimationImputationLanguage ModelingLanguage Modelling+4

Masked Latent Semantic Modeling: an Efficient Pre-training Alternative to Masked Language Modeling

2023-07-07 · ACL Findings 2023 7 · Gábor Berend

In this paper, we propose an alternative to the classic masked language modeling (MLM) pre-training paradigm, where the objective is altered from the reconstruction of the exact identity of randomly selected masked subwo…

Language ModelingLanguage ModellingMasked Language Modeling

Exemplar Retrieval Without Overhypothesis Induction: Limits of Distributional Sequence Learning in Early Word Learning

2026-04-06 · Jon-Paul Cacioli arxiv

Background: Children do not simply learn that balls are round and blocks are square. They learn that shape is the kind of feature that tends to define object categories -- a second-order generalisation known as an overhy…